BrainxBot — Pankaj Singh · AI Trading Automation

The market is a black box. So is the bot.

You don't need to predict the future. You need to recognise when the present looks like a past pattern that paid.

BrainxBot is built on a single conviction: markets leak. They leave fingerprints in price action, volume profiles, volatility clusters, and order-flow rhythms — patterns that repeat just often enough to be edge.

The bot ingests live market data, scores it against thousands of historical pattern signatures, and executes only when statistical confidence crosses threshold. No emotion. No FOMO. No revenge trades. Just signal, position, exit.

"The efficient market hypothesis is correct in a few aspects, but not in others… we identify anomalies in the data. We've built models that predict prices well enough to make money."

— Jim Simons, Renaissance Technologies

This is the black-box school of quantitative trading — the same philosophy that took Simons' Medallion Fund to roughly 66% annualised gross returns over three decades. The model doesn't need to explain itself. It needs to be right more than it's wrong, and disciplined enough to act on it.

PHILOSOPHY · LOG 001

Pattern is memory.
Memory is edge.

Three principles the system never violates: trade only when the pattern is confirmed, size only what the volatility allows, and never override the model in the middle of a thesis.

24/7

Market scan

0ms

Emotion lag

∞

Patterns logged

1×

Plan, executed

Four stages. One loop, forever.

STAGE 01

Ingest

Live price, volume, volatility, and order-flow data streams in from your broker's API — second by second, across every active pair.

STAGE 02

Recognise

The pattern engine scores the current market state against historical signatures. Only setups above the statistical threshold pass through.

STAGE 03

Execute

Position size is derived from volatility and risk per trade. Orders fire directly into your broker account — entries, stops, and targets in one motion.

STAGE 04

Learn

Every closed trade — winner or loser — is logged and fed back. The model's view of what works updates with every cycle. The edge compounds.

Pankaj Singh, in his own words.

NamePankaj Singh

RoleFounder, Operator

BasedIndia

DomainQuant · Automation

CHAPTER 01 · THE OBSESSION

A retail trader who got tired of being wrong.

I started where most people start — on a phone, with a chart, with a gut feeling, and with the slow realisation that gut feelings are expensive. The market took its tuition. I started keeping logs of every trade — what I felt, what I did, what actually happened. The gap between the three was the problem.

CHAPTER 02 · THE PIVOT

Reading Simons, reading the data.

The Medallion Fund's story rewired how I saw the markets. Renaissance didn't predict — they found patterns, measured edge, and let mathematics do the heavy lifting. I stopped trying to be a genius and started trying to be systematic. Every "intuition" I had, I tried to encode as a rule. Every rule, I tried to backtest.

CHAPTER 03 · THE BUILD

BrainxBot, version one.

The first version was ugly — a script, a few API keys, a spreadsheet. But it worked. It removed me from the decision loop, which turned out to be the single biggest source of leakage in my P&L. Version after version, the bot got smarter, the pattern library got deeper, and the risk model got tighter.

CHAPTER 04 · THE SERVICE

Running it for other accounts.

People started asking. The model now runs on multiple client accounts across Delta Exchange India, Binance, and Exness. The arrangement is simple — you keep your funds, I deploy the bot, the bot trades, you watch. No black-box bank account. Just a black-box strategy on your white-box exchange.

CHAPTER 05 · WHAT'S NEXT

More patterns. More markets. Same discipline.

The roadmap is more of the same, just deeper: more pattern signatures, more asset classes, tighter feedback loops, better drawdown defence. The principle never changes — trade the patterns, not the noise, and never override the model.

Private — Unlocked

Dreambook

Notes to self. Not for the timeline.

CHAPTER · I

The dream.

A bot that lets a hundred people sleep at night, while it does the trading they would have done worse, awake.

Not a fund. Not a guru program. Not a Telegram channel. Just a quiet, disciplined piece of code that runs on each client's own account, asks nothing of them but the chart to read, and never loses more in a day than it's allowed to.

Ten years from now, the dream is to have stopped marketing. The track record will market itself, every account a quiet referral.

CHAPTER · II

The numbers.

Private targets. Not for the website footer. Honest measures of progress.

— —

AUM, end of year

— %

Monthly target

—

Client accounts

— %

Max drawdown allowed

Fill these in. Update them quarterly. Don't lie to yourself.

CHAPTER · III

The lessons.

Earned the hard way. Each one cost real money.

The model has no ego. You do. Get out of the way.
The trade you almost took, but didn't, doesn't exist. Stop counting it.
A losing day is a sample. A losing week is a data point. A losing month is a question.
Position size is not a feeling. It's a number that comes from volatility.
The market doesn't owe you a win to make you whole. Move on.
If you can't explain why the bot took the trade, you can't override it.

CHAPTER · IV

The reminders.

Trade the patterns, not the noise.

— Mantra one

The model doesn't need to explain itself. It needs to be right more than it's wrong.

— Mantra two

Discipline isn't what you do on a winning day. It's what you do on a losing one.

— Mantra three

CHAPTER · V

The mentors.

People I've never met but learn from every week.

Jim Simons — Renaissance Technologies. Proved the black-box school works.
Ed Thorp — Beat the dealer, then beat the market. The original quant.
Nassim Taleb — On fragility, tail risk, and not being a fool.
Naval Ravikant — On compounding, patience, and the long game.

Trade the
patterns, not
the noise.

The market is a black box. So is the bot.

Pattern is memory.
Memory is edge.

Data goes in. Mystery happens.
Probability comes out.

Four stages. One loop, forever.

Ingest

Recognise

Execute

Learn

Your account. Your custody. Your control.

Pankaj Singh, in his own words.

A retail trader who got tired of being wrong.

Reading Simons, reading the data.

BrainxBot, version one.

Running it for other accounts.

More patterns. More markets. Same discipline.

Dreambook

Dreambook

The dream.

The numbers.

The lessons.

The reminders.

The mentors.

Run the
bot on your
account.

Trade the patterns, not the noise.

The market is a black box. So is the bot.

Pattern is memory.Memory is edge.

Data goes in. Mystery happens.Probability comes out.

Four stages. One loop, forever.

Ingest

Recognise

Execute

Learn

Your account. Your custody. Your control.

Pankaj Singh, in his own words.

A retail trader who got tired of being wrong.

Reading Simons, reading the data.

BrainxBot, version one.

Running it for other accounts.

More patterns. More markets. Same discipline.

Dreambook

Dreambook

The dream.

The numbers.

The lessons.

The reminders.

The mentors.

Run thebot on youraccount.

Trade the
patterns, not
the noise.

Pattern is memory.
Memory is edge.

Data goes in. Mystery happens.
Probability comes out.

Run the
bot on your
account.